- Resource Center
- Professional Development
- Articles & Videos
- How to Create a Successful Localization Process in 2023
4 July 2023
How to Create a Successful Localization Process in 2023
Sign up for our newsletter on globalization and localization matters.
In today’s globally connected world, localization is crucial to a company’s expansion strategy. And yet, businesses setting up their first localization process – or overhauling their existing approach – are quickly overwhelmed by an overabundance of information, tools, and available language service providers (such as content translation agencies).
Companies without dedicated translation resources, such as an in-house translation manager or a translation department, can be especially overwhelmed by the number of different services they need to explore and choose from.
In this article, we take a step back from investigating all the nuances of localization. Instead, we define the three key elements of a robust, scalable, and quality-focused translation strategy.
1. Find the right people
Let’s quickly look at recent developments. Large language models, or neural networks trained on vast amounts of text (such as ChatGPT), are more sophisticated than ever. And classical neural machine translation engines (NMTs), such as Google Translate, deliver generally good results.
The outcome of these developments? Companies are more likely to rely solely on these tools for localization. After all, pure machine translation is quicker, cheaper, and more scalable than any localization workflow involving a human (or humans).
But, both NMTs and large language models (LLMs) still can’t provide the same quality as human translators: NMTs struggle to grasp linguistic nuances conveyed through text, such as irony, jokes, and puns. While LLMs don’t deliver consistent translations and are prone to hallucinations (randomly adding or removing content) and concordance errors.
What does all this mean? Finding the right people for a company’s localization needs – especially for external content such as a homepage, marketing materials, and social media posts – is still a crucial part of a localization strategy.
And there are many choices: will the business use full-time in-house translators, hand-picked freelancers, or a mix of in-house and freelance translators from a trusted LSP? That depends on a company’s goals, needs, budget, internal expertise, and other factors.
A robust, scalable localization setup doesn't just involve translators – there are a multitude of other roles involved in the process as well:
● Translation manager(s)
● Reviewers
● Desktop publishing specialists
● Linguistic engineers
● Tech delivery specialists
● Quality managers
● Vendor managers
While businesses can fill some of these roles using in-house experts within, say, design or marketing departments, choosing the right people is always a challenge. That means, sooner or later, most companies will begin procuring one or even multiple LSPs.
Thorough test translations with industry-specific, moderately challenging texts are a great way to assess an LSP’s level of quality and see how they navigate common issues such as queries and unclear terminology.
But throughout the test translation process, it’s worth remembering that the LSPs will use their best resources and as much quality assurance as possible to win a new client. So, bear in mind that the translation quality might drop slightly after LSP onboarding.
Once a company’s localization needs are defined and the right people for the job are tested and onboarded, the translation process can start. But translation projects tend to become complex quickly just by the sheer number of people involved: simultaneously translating 10 pieces of copy into 10 different languages can include more than 200 individual contributors!
At this point, efficiency and standardization are critical. So, next we explore another vital ingredient of a successful localization strategy.
2. Choose the best tools and technology
Now that all localization stakeholders are onboarded and ready, it’s time to explore the translation process’s environment and technology. Currently, the trend is to use cloud-based, all-in-one solutions.
So, here we explore the benefits of these solutions and how they fit into the overall translation process.
Considering that professional tools are usually quite expensive, it’s natural to question whether a company needs to make such an investment. Here, the answer isn’t a clear yes or no.
Why? Well, the ‘yes’ part is that everybody working on translations, from translation managers to freelance translators, will need proper tooling to do the best job possible.
But the ‘no’ part of our answer is that only some companies have to procure their own tools. Other businesses can easily use their LSP’s setup – or even no tools at all – if they share their translatable files with their LSP in another way.
It’s also worth remembering that businesses usually can’t outsource all translations due to urgency, confidentiality, or cost-saving reasons. If that’s the case, having an in-house translation tool that is compatible with the LSP’s setup is highly advantageous for the following reasons:
● Automated workflows and file transfer
● Shared translation assets (Translation Memories, Termbases, RegEx files, etc.)
● Less quality drift between in-house and LSP’s translations
● Better control and maintenance of company-owned translation assets
● More data and insights for internal reporting and decision making
● More flexibility with LSP on- and off-boarding (in a worst-case scenario)
Businesses looking for the right in-house translation tools have many options. Most modern cloud-based tools are very similar in setup, functionalities, and pricing. Yet, two main points can heavily influence tool selection:
1. Tool suitability: Nowadays, there’s more use-case tailored translation tools available than ever, especially UI/Software localization tools that are hard to use for longer content. So, understanding the relative shares of all translatable content types within a company is critical during the tool selection process.
2. Content connectors: Similar to the previous point, different tools also offer content connectors that seamlessly integrate and exchange files with other tech, including Google Drive, WordPress, GitHub, Salesforce, and many others. While almost any tool offers the most popular connectors, more specialized connectors might be the differentiating factor when choosing the right translation tool.
Choosing the right tool requires the same amount of due diligence as selecting a LSP. After all, a low-quality or incorrectly-used tool can significantly undermine an entire company’s trust in the overall localization strategy.
But how can a business ensure that its chosen tools and the people involved in the translation strategy produce the best work possible? Let’s take a look.
3. Develop clear instructions and processes
The third crucial aspect of a robust localization workflow is clearly defined processes and instructions. Even though a business can (and probably will) constantly refine and adjust them on the go, clearly defined processes result in fewer disruptions and difficulties as they scale the translation process.
But what do these instructions and processes look like?
The first thing a company can do before procuring translation provider(s) or tools of any sort is to create two types of documents: style guides and glossaries.
An overall, language-agnostic style guide in English is usually sufficient to start building a good workflow, while a diligent LSP will provide support to create and refine language-specific versions later on.
Unlike a glossary, a set of style guides does require a certain amount of cross-functional effort in the early phase but is much easier to maintain in the long term, as it will more or less settle within the first one or two years of active use.
Glossaries, or just one company-wide glossary, are relatively straightforward to set up. However, they usually need more continuous effort to keep them updated. This is because the terminology can quickly change, and if actively used, glossaries become polluted with a surplus of generic terms over time. So, it’s a good idea for businesses to run periodic glossary reviews with in-house subject matter experts at least once a year.
Compared to style guides and glossaries, clear localization processes are much more difficult to define by a company with no in-house translation expertise. A standard, easy workflow (copywriting - translation - review - publishing) can quickly grow to involve 10-15 individual (although partially automatable) steps, especially when localizing a wide range of assets. And every company will require multiple, slightly deviating workflows for different content types.
Here, an experienced translation manager (either in-house or on the LSP’s side) can provide essential support. In close collaboration with various company stakeholders, they will be able to identify levels of quality and the corresponding tools and techniques required for different content types. A translation manager will also play a crucial role in ensuring that all steps and the reasoning behind them are fully understood by all internal and external stakeholders and followed correctly by each contributor.
A recipe for perfect (and simplified) translations
Translations are a highly complex topic that requires a lot of expertise, from the first sketches of a localisation workflow to the delivery of every translatable file. But, with the right people armed with the right processes and tools, any company can provide its customers with top-notch content in any language it needs.
And, if things don’t work out at first, always remember: it’s processes, not people, that fail.
Do you want to contribute with an article, a blog post or a webinar?
We’re always on the lookout for informative, useful and well-researched content relative to our industry.
Ilya Urzhumov
Global Translations Coordinator at Mollie, one of Europe's fastest growing fintechs. State-certified EN-DE translator (specializations: technical, scientific and medical translations). Former Lead German Translator for one of RWS' Top 3 largest clients with a yearly throughput of several million words per language.